|
Cost-effectiveness of alternative treatments for women with osteoporosis in Canada |
Goeree R, Blackhouse G, Adachi J |
|
|
Record Status This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn. Health technology Four pharmacological therapies for women with osteoporosis were examined. These were three bisphosphonates (etidronate, alendronate and risedronate) and raloxifene. Raloxifene was assumed to be given at a dose of 60 mg/day, alendronate at 70 mg once weekly, and risedronate at 35 mg once weekly. Patients on etidronate were assumed to be taking a regimen pack of 400 mg for 14 days, followed by 500 mg of elemental calcium (as calcium carbonate) for 76 days. All therapies were assumed to be given for a maximum of 5 years.
Type of intervention Treatment and secondary prevention.
Economic study type Cost-effectiveness analysis and cost-utility analysis.
Study population The study population comprised a hypothetical cohort of post-menopausal women with osteoporosis, with a mean age of 65 years, who did not have a prior vertebral fracture.
Setting The setting was secondary care. The economic study was carried out in Canada.
Dates to which data relate The effectiveness data were derived from studies published between 1986 and 2005. The costs and resource use data were derived from studies published between 1993 and 2005. The price year was 2005.
Source of effectiveness data The effectiveness data were derived from a synthesis of published studies and authors' opinions.
Modelling A Markov model was constructed to evaluate the long-term costs and benefits of the alternative treatments for osteoporosis. The time horizon of the model was 30 years, meaning that the women were followed until the age of 95 years or death. Yearly cycles were considered. The health states of the model were healthy, hip fracture (first and subsequent years), vertebral fracture (first and subsequent years), BC (first and subsequent years), CHD (first and subsequent years), VTE in the first year, and a combination of health states. The CHD health states were divided into further health states such as recognised acute myocardial infarction (AMI), unrecognised AMI, angina pectoris, coronary insufficiency, and sudden death. Every year, women could remain in the same health state, move into one of the other health states, or die from "other causes". A simplified structure of the model was represented graphically.
Outcomes assessed in the review The outcomes assessed were:
the risk of disease events included in the decision model,
the mortality risk,
the effect of treatment, and
the utility weights associated with specific health conditions.
Study designs and other criteria for inclusion in the review The authors did not describe a systematic review of the literature. It was stated that, whenever possible, systematic literature reviews, randomised controlled trials (RCTs), or meta-analyses of trials were used to derive clinical data on the treatment effect of the four therapies compared. Patient characteristics at baseline were similar to those included in an RCT of raloxifene. Data on relative risks of events (fractures and other health conditions) for osteoporotic women compared with the general population came from different sources: the Canadian Institute of Health Information's hospital discharge abstract database for the provinces of Ontario, British Columbia and Alberta; a population-based study from Rochester (New York); the US Department of Health and Human Service's Surveillance, Epidemiology, and End Results; and Canadian Life Tables. The utility values related to osteoporosis were derived from the Canadian Multicenter Osteoporosis Study, a large multi-centre study. The utility weights for other health conditions were obtained from published studies, but no details were given.
Sources searched to identify primary studies Criteria used to ensure the validity of primary studies The authors stated that, for treatment effect, high-level evidence (RCTs, meta-analysis of RCTs) was included whenever available.
Methods used to judge relevance and validity, and for extracting data Number of primary studies included Twenty-three primary studies provided the clinical data.
Methods of combining primary studies Investigation of differences between primary studies Results of the review The risks of disease events (hip fractures, vertebral fractures, CHD, BC and VTE) for the general population were age-dependent and were not explicitly reported.
The increased risk of vertebral fracture incidence among women with osteoporosis (without previous fractures) was 2. The increased risk of fracture incidence among women with established osteoporosis (i.e. with previous fractures) was 1.72 for hip fractures and 3.81 for vertebral fractures.
The mortality rates for the general population and women with osteoporosis were not reported (age-dependent).
For treatment effect, the relative risk (RR) of hip fracture was 0.68 (95% confidence interval, CI: 0.30 to 1.54) for alendronate and 1 (no impact) for the other three treatments. The RR of vertebral fractures was 0.60 (95% CI: 0.46 to 0.80) for alendronate, 0.40 (95% CI: 0.20 to 0.83) for etidronate, 0.53 (95% CI: 0.24 to 1.17) for risedronate, and 0.53 (95% CI: 0.32 to 0.88) for raloxifene.
Etidronate, alendronate and risedronate had no impact on the RR of CHD, while raloxifene had a non significant impact on CHD in all patients (RR 0.82, 95% CI: 0.56 to 1.22) and a non significant impact on CHD in high-risk patients (RR 0.66, 95% CI: 0.37 to 1.19).
Etidronate, alendronate and risedronate had no impact on the RR of BC, but raloxifene had an RR of BC of 0.38 (95% CI: 0.24 to 0.58).
Etidronate, alendronate and risedronate had no impact on the RR of VTE, but raloxifene had an increased risk of VTE of 2.35 (95% CI: 1.22 to 4.52).
The utility estimate was 0.85 for osteoporotic women aged 50 to 64 years, 0.81 for osteoporotic women aged 65 to 75 years, and 0.71 for osteoporotic women aged 75 years and older.
Other utility values were not reported.
Methods used to derive estimates of effectiveness The authors made some assumptions that were used in the decision model.
Estimates of effectiveness and key assumptions It was assumed that treatment continued for 5 years and that women were fully compliant with therapy. Excess mortality due to VTE occurred within 1 week. Mortality for hip fracture, vertebral fracture and VTE patients in subsequent years was assumed to be at the same rate as that for "other causes" of death.
Measure of benefits used in the economic analysis The summary benefits used in the economic analysis were the life-years (LYs) and quality-adjusted life-years (QALYs). These were estimated using the modelling approach. The QALYs were derived by combining utility estimates and survival obtained from the review of the literature. Future benefits were discounted at an annual rate of 5%.
Direct costs The analysis of the costs was carried out from the perspective of the third-party payer. It included the costs associated with the medications under examination and the treatment of health conditions such as hip and vertebral fractures, BC, CHD and VTE. The costs of medication included pharmacy mark-up charges, dispensing fees, specialist visits, blood work and dexa scan. The unit costs and the quantities of resources used were generally not presented separately but as macro-categories. The unit costs and quantities of resource use were reported only for a very few items, such as professional fees and tests. The drug costs were derived from the Ontario Drug Benefit plan, while professional fees were derived from the Ontario Schedule of Benefits. The costs of tests came from a hospital participating in the Ontario Case Costing Project. The costs of health conditions were evaluated separately for the first year and for subsequent years. These costs came from different studies and were often age-dependent. Data on resource consumption were presumably derived from the same studies as those used to derive the costs. Discounting was relevant, as the long-term costs were considered, and an annual discount rate of 5% was used. The price year was 2005. Where necessary, the costs were updated to 2005 values using the Consumer Price Index.
Statistical analysis of costs The costs were treated deterministically in the base-case.
Indirect Costs The indirect costs were not relevant from the perspective of the third-party payer and were therefore not included.
Sensitivity analysis A probabilistic sensitivity analysis was run by assigning stochastic distributions to most model inputs. One thousand 1,000 Monte Carlo simulations were conducted. Specifically, log-normal distributions were assigned to underlying the risk and RR reductions for each disease, gamma distributions were used for the annual costs of treating each health condition, and beta distributions were assigned for utility values. Cost-effectiveness acceptability curves for different ceiling ratios (willingness-to-pay levels) were generated. Univariate sensitivity analyses were also carried out to evaluate the robustness of the base-case cost-utility ratios to variations in selected model assumptions and items such as treatment duration, time horizon, discount rate, compliance, impact of raloxifene on BC and CHD, risk of fracture, starting age, elevated CHD risk, women with established osteoporosis, and vertebral facture risk. Finally, a threshold analysis was performed to determine at what price drug treatments would become cost-effective. Alternative values were mainly derived from the literature or from authors' assumptions.
Estimated benefits used in the economic analysis The expected LYs were 11.541 with no intervention, 11.555 with etidronate, 11.569 with alendronate, 11.569 with raloxifene and 11.552 with risedronate.
The expected QALYs were 8.462 with no intervention, 8.523 with etidronate, 8.525 with alendronate, 8.541 with raloxifene and 8.508 with risedronate.
Cost results The total costs (intervention plus disease costs) were CAD 15,960 with no intervention, CAD 17,940 with etidronate, CAD 18,016 with alendronate, CAD 19,841 with raloxifene and CAD 19,538 with risedronate.
Synthesis of costs and benefits Incremental cost-effectiveness ratios and cost-utility ratios were calculated to combine the costs and benefits of the alternative strategies.
In comparison with no intervention, the incremental cost per LY gained was CAD 143,247 with etidronate, CAD 72,883 with alendronate, CAD 140,782 with raloxifene and CAD 346,872 with risedronate.
In comparison with no intervention, the incremental cost per QALY gained was CAD 32,571 with etidronate, CAD 32,760 with alendronate, CAD 49,279 with raloxifene and CAD 78,275 with risedronate.
When compared with the next most effective strategy, the incremental cost per QALY was CAD 32,571 with etidronate over no intervention, CAD 38,623 with alendronate relative to etidronate, and CAD 114,070 with raloxifene over alendronate. Risedronate was dominated by both etidronate and alendronate.
The probabilistic sensitivity analysis showed that, below a willingness-to-pay per QALY of CAD 32,000, no intervention had the highest probability of being the most cost-effective strategy. For ceiling ratios between CAD 32,000 and CAD 116,000, alendronate had the highest probability of being the most cost-effective strategy. For a threshold above CAD 116,000, raloxifene had the highest probability of being the most cost-effective treatment.
The deterministic sensitivity analysis suggested some interesting results. For example, a longer treatment duration would improve the cost-effectiveness of etidronate and would make alendronate and raloxifene less attractive. If the duration of the model decreased, the relative cost-effectiveness of all active treatments decreased, especially for alendronate. The relative cost-effectiveness of all active treatments decreased also with lower compliance and when the set time for the risk of fracture once therapy was discontinued decreased. The use of generic equivalents would favour the most expensive treatments (risedronate and raloxifene).
When patient sub-groups and patients with different risk factors were considered, the sensitivity analysis showed that if therapy started at age 55 years, then raloxifene became relatively more cost-effective and alendronate was dominated by etidronate or raloxifene. If therapy were initiated at age 75 years, alendronate became much more attractive and dominated both etidronate and raloxifene. In patients with elevated CHD risk, raloxifene was more cost-effective, even at younger starting ages, while alendronate was dominated by etidronate and raloxifene. Finally, for women with established osteoporosis, alendronate was cost-effective while etidronate and raloxifene were dominated.
Authors' conclusions In the base-case analysis of a 65-year-old woman with osteoporosis but no previous vertebral fracture, receiving therapy for 5 years, etidronate and alendronate were the most cost-effective treatments using a threshold of CAD 50,000 per quality-adjusted life-year (QALY) gained in Canada. The sensitivity analysis suggested that raloxifene was more cost-effective for younger women, as well as for women with an elevated risk of coronary heart disease (CHD). However, all active treatments were generally cost-effective compared with no intervention when using standard thresholds for willingness-to-pay.
CRD COMMENTARY - Selection of comparators The authors did not provide an explicit justification for the choice of the comparators, but they are the most commonly used pharmacological treatments for osteoporosis. Dosages were reported and alternative treatment durations were considered. You should decide whether they are valid comparators in your own setting.
Validity of estimate of measure of effectiveness The effectiveness data came from a review of the literature, the methods and conduct of which were not reported. Few details of the inclusion criteria and characteristics of the primary studies were given. The authors stated that, whenever possible, high-quality evidence was selected. Treatment effect was therefore taken mostly from RCTs or meta-analyses of trials. However, the issue of heterogeneity across the primary sources was not addressed and it was unclear whether treatment effect was based on direct or indirect comparisons across treatments. Although the use of assumptions introduces further uncertainty into the clinical estimates used in the model, the authors performed extensive sensitivity analysis (both probabilistic and deterministic) to consider different scenarios and alternative base-case values.
Validity of estimate of measure of benefit LYs and QALYs are appropriate benefit measures, not only because they capture the most relevant dimensions of health (survival and quality of life) for women with osteoporosis but also in terms of their relative ease of comparison with the benefits of other health care interventions. Discounting was performed in accordance with international guidelines. Some information on the sources of utility values was provided for the case of osteoporosis (Canadian values) but not for other health conditions.
Validity of estimate of costs The analysis of the costs was consistent with the perspective of the economic analysis. The costs were presented as macro-categories and, in general, a detailed breakdown of cost items was not provided. This limits the possibility of replicating the analysis in other settings. The sources of the costs were not described for all cost categories, particularly the quantities of resources used. The cost estimates were treated deterministically in the base-case but stochastically in the sensitivity analysis. The authors considered mark-up charges and dispensing fees in the calculation of the drug costs. The sensitivity analysis also investigated the impact of generic equivalents for some drugs. The price year was reported, which will facilitate reflation exercises in other time periods.
Other issues The authors reported the results from other economic evaluations and these were similar to those achieved in the current study. The issue of generalisability of the study results to other settings was addressed in the sensitivity analysis, which involved an extensive consideration of the alternative scenarios. The authors noted that their analysis had the usual drawbacks of modelling studies (i.e. complexity and use of assumptions), although the current model was validated in the Canadian setting. The use of population-based studies and Canadian data for costs and some quality of life values represents a strong characteristic of the analysis. However, even when data from other countries were used, the best available evidence was selected. A caveat of the analysis was the fact that gastrointestinal disturbances that might be associated with the use of bisphosphonates were not modelled.
Implications of the study The study results suggest that using conventionally quoted thresholds for cost-effectiveness and in comparison with no intervention, alendronate, etidronate and raloxifene were all cost-effective strategies for post-menopausal women with osteoporosis but no previous vertebral fracture. The relative cost-effectiveness among these therapies depends on patient characteristics and assumptions about treatment duration.
Source of funding Initial funding from Eli Lilly Canada.
Bibliographic details Goeree R, Blackhouse G, Adachi J. Cost-effectiveness of alternative treatments for women with osteoporosis in Canada. Current Medical Research and Opinion 2006; 22(7): 1425-1436 Other publications of related interest Because readers are likely to encounter and assess individual publications, NHS EED abstracts reflect the original publication as it is written, as a stand-alone paper. Where NHS EED abstractors are able to identify positively that a publication is significantly linked to or informed by other publications, these will be referenced in the text of the abstract and their bibliographic details recorded here for information.
Stevenson M, Jones ML, De Nigris E, et al. A systematic review and economic evaluation of alendronate, etindronate, risendronate, raloxifene and teriparatide for the prevention and treatment of postmenopausal osteoporosis. Health Technol Assess 2005;22(1).
Kanis JA, Brazier JE, Stevenson M, et al. Treatment of established osteoporosis: a systematic review and cost-utility analysis. Health Technol Assess 2002;6(29).
Kanis JA, Borgstrom F, Johnell O, et al. Cost-effectiveness of raloxifene in the UK: an economic evaluation based on the MORE study. Osteoporosis Int 2005;16:15-25
Indexing Status Subject indexing assigned by NLM MeSH Aged; Alendronate /economics /therapeutic use; Bone Density Conservation Agents /economics /therapeutic use; Canada; Cost-Benefit Analysis; Decision Support Techniques; Etidronic Acid /analogs & Female; Humans; Markov Chains; Osteoporosis, Postmenopausal /drug therapy /economics; Raloxifene Hydrochloride /economics /therapeutic use; Risedronate Sodium; derivatives /economics /therapeutic use AccessionNumber 22006001448 Date bibliographic record published 31/05/2007 Date abstract record published 31/05/2007 |
|
|
|